Factors Contributing to Dwarfing in the Mangrove Avicennia marina

Abstract

•Background and Aims In Richards Bay, South Africa, Avicennia marina frequently exhibits a distinct productivity gradient, with tree height decreasing markedly from 6–10 m in the fringe zone to <1·5 m in the dwarf zone which is 120 m inland at a slightly higher elevation. In this investigation, soil physico-chemical conditions between fringe and dwarf A. marina were compared and the constraints imposed by any differences on mangrove ecophysiology and productivity determined.

•Methods Soil and plant samples were analysed for inorganic ions using spectrophotometry. Gas exchange measurements were taken with an infrared gas analyser and chlorophyll fluorescence with a fluorometer. Xylem ψ was determined with a pressure chamber and chlorophyll content with a chlorophyll absorbance meter.

•Results In the dwarf site, soil salinity, total cations, electrical conductivity and soil concentrations of Na⁺, K⁺, Ca²⁺, Mg²⁺, Zn²⁺, Mn²⁺ and Cu²⁺ were significantly higher than those in the fringe zone. Soil water potential and the concentration of soil P, however, were significantly lower in the dwarf site. In the leaves, Na⁺ was the predominant ion and its concentration was 24 % higher in dwarf than fringe mangroves. Leaf concentrations of K⁺, Ca²⁺, Mg²⁺, Mn²⁺ and P, however, were significantly lower in dwarf mangroves. Photosynthetic performance, measured by gas exchange and chlorophyll fluorescence, was significantly reduced in the dwarf plants.

•Conclusions The results suggest that hydro-edaphic factors contribute to high soil salinities, low water potentials, water stress and ion imbalance within tissues including P deficiency, which in interaction, contribute to dwarfing in Avicennia marina.

INTRODUCTION

Mangroves constitute an important ecosystem because of their global extent and high productivity (Mitsch and Gosselink, 2000). These plants thrive in intertidal zones of the tropics and subtropics that are characterized by regular tidal inundation and fluctuating salinity. Mangrove species are well adapted, both morphologically and physiologically, to survive under saline conditions (Ball, 1996; Sobrado and Ball, 1999; Naidoo et al., 2002). Despite their ecological success in saline environments, however, carbon assimilation capacity and growth are reduced as salinity increases (Ball and Farquhar, 1984; Sobrado and Ball, 1999; Naidoo and Chirkoot, 2004).

In Richards Bay, Avicennia marina is the dominant mangrove, often occurring in monospecific stands over a wide range of salinities. In some localities, natural productivity gradients occur over a relatively short distance. Tree height decreases markedly from 6–10 m in the fringe zone to <1·5 m in the dwarf zone which is 120 m inland at a slightly higher elevation. Dwarfing in mangroves has been attributed to a variety of factors including high salinity, poor aeration (Davies, 1940), waterlogging and salinity (Egler, 1952), compacted peat (Craighead, 1971) and nutrient limitation (Lugo and Snedakar, 1974; Davis et al., 2001; Lovelock et al., 2004). Recent evidence suggests that phosphorus availability may be an important factor in dwarf mangrove production, especially in carbonate-dominated environments (Feller, 1995; Davis et al., 2001; McKee et al., 2002; Lovelock et al., 2004). The majority of these studies were undertaken on dwarf Rhizophora mangle mangroves in the Florida Everglades or in the Caribbean. There is a paucity of data on edaphic conditions that contribute to dwarfing in other mangroves species in other parts of the world, and on ecological processes that operate under these conditions.

In this investigation soil physico-chemical conditions, water and ion relationships and photosynthetic performance of fringe and dwarf A. marina were compared. The objectives of this study were to (a) determine differences in soil physico-chemical conditions between sites with fringe and dwarf mangroves, (b) investigate how these differences affect ion and water relationships, and (c) determine constraints imposed by these environmental conditions on mangrove photosynthesis.

MATERIALS AND METHODS

Study site

This study was undertaken in Richards Bay Harbour (28°48′S, 32°05′E), South Africa, in a monospecific Avicennia marina (Forsk.) Vierh. stand. In the selected site, a natural productivity gradient occurred from the fringe to the dwarf site, 120 m inland at a slightly higher elevation. The fringe zone, which is regularly inundated twice daily by tides, supports luxuriant adult A. marina trees that are 6–10 m tall and which form a dense, well-developed canopy. The landward zone which is only inundated during high spring tides, supports diminutive or dwarf A. marina trees that are <1·5 m in height. These trees are sparse and in poor condition. Adjacent to the dwarf A. marina is a barren zone with hardly any vegetation, except for isolated Sarcocornia natalenis and Salicornia spp. Two sites with A. marina were selected for measurements, one in the fringe zone and the other in the dwarf zone.

Soil analyses

Soils were sampled twice in September 2003, twice in December 2003 and three times in June 2004 during three spring and four neap tides. Samples were obtained randomly from each site at a depth of 10–20 cm. Soils were air-dried, crushed with a wooden mallet, passed through a 2-mm sieve and analysed for inorganic ions using the procedures described by Hunter (1975) and Farina (1981). Exchangeable Na⁺ was extracted with 1 m ammonium acetate, Ca²⁺ and Mg²⁺ by 1 m KCl, and K⁺, P, Zn²⁺, Cu²⁺ and Mg²⁺ by the Ambic-2 extracting solution (Hunter, 1975). Na⁺, K⁺, Ca²⁺, Mg²⁺, Zn²⁺, Cu²⁺ and Mn²⁺ were determined by atomic absorption (Varian Spectra AA-10, Mulgrave, Australia). P was determined by the molybdenum blue procedure (Hunter, 1975) and a spectrophotometer (Beckman DU 640, Fullerton, USA). Saturation extracts of soils were analysed for electrical conductivity using an electrical conductivity meter. Soil pH was determined on air-dried samples using a 1 : 2·5 soil to 1 m KCl ratio (Hunter, 1975). The suspension was stirred for 5 min and allowed to stand for 30 min. The pH was measured with a glass combination electrode while stirring. Bulk density was determined using a plastic cylinder to take a sample of known volume that was then oven-dried at 105 °C and dry mass determined.

Plant analyses

Leaves were collected for ion analyses twice in September 2003, twice in December 2003 and once in June 2004 during two spring and three neap tides. Fifty fully expanded mature leaves were removed from each of five representative trees at each site, rinsed for 10 s in distilled water to remove surface salt and dried at 70 °C to constant mass. Samples were milled through a 1-mm screen and stored in plastic vials. Subsamples were dry-ashed at 450 °C and dissolved in 1 m HCl. Concentrations of Na⁺, K⁺, Ca²⁺, Mg²⁺, Cu²⁺, Mn²⁺ and Zn²⁺ were determined by atomic absorption (Varian Spectra AA-10, Mulgrave, Australia), P by the molybdenum blue procedure, and N by the automated Dumas dry combustion method, using a LECO CNS 2000, Leco Corporation, Michigan, USA (Matejovic, 1996). All chemical analyses were verified by the use of tissue standards.

CO₂ exchange

Gas exchange measurements were taken in September and December 2003 and January, April, June and August 2004, during three spring and three neap tides. Measurements were taken with a portable, infrared gas analyser (Licor-6400; Licor, Lincoln, NB, USA) under ambient conditions. All measurements were taken on mature, north-facing, fully exposed, sun leaves. The leaf chamber was clamped onto the leaf and tilted to maximize light incident on the leaf. Generally, leaves of A. marina are inclined at an angle of 0–45° from the horizontal (Tuffers et al., 1999). All measurements were taken between 1000 h and 1200 h at saturating light (1200–1500 μmol m^–2 s^–1) and ambient temperature (28–30 °C). On each of the six measurement days, five representative trees were selected and six measurements made per tree, resulting in a total of 180 measurements per site. On each day measurements were taken on different trees.

Chlorophyll fluorescence

Chlorophyll fluorescence was determined with a field portable, pulse amplitude, modulated fluorometer (PAM-2100, Walz, Effeltrich, Germany). Fluorometer operation and data processing were conducted with a Hewlett Packard palmtop computer (HP 200LX). All measurements were taken on the lamina, midway between the base and tip of mature leaves. Quantum yield of photosystem II (PSII) electron transport (ΔF/F_m′) was calculated as (F_m′ – F)/F_m′ (Genty et al., 1989) where F is the light-adapted fluorescence when a saturating light pulse of 7500 μmol m^–2 s^–1 PPFD for 700 ms duration is superimposed on the prevailing environmental irradiance level (Schreiber et al., 1995). Electron transport rate (ETR) through PSII was calculated as 0·5 × 0·84 × PPFD × ΔF/F_m′ assuming that 84 % of incidental light is absorbed by leaves and that photons are equally distributed between PSII and PSI (Schreiber et al., 1995). Measurements of chlorophyll fluorescence were taken under ambient conditions at saturating light on the same day and on similar leaves on which gas exchange measurements were made. On each of six days, five representative trees were selected at each site and six measurements taken per tree. Different trees were randomly selected on each of the six measurement days. The intrinsic efficiency of light energy conversion of PSII (F_v/F_m) was measured after 30-min dark adaptation with a dark leaf clip (Walz, Effeltrich, Germany). For F_v/F_m, five measurements were taken on each tree over the six measurement days as described for the other fluorescence measurements.

Water potential (ψ)

Xylem ψ was determined in September and December 2003 and June 2004 during neap, neap and spring tides, respectively. Three measurements were taken per site on each day at midday with a Scholander pressure chamber (Soil Moisture Equipment Corporation, Santa Barbara, CA, USA). Soil ψ was determined on soil samples obtained at 10 cm depth from the fringe, dwarf and barren sites with a WP4 Dewpoint Potential Meter (Decagon Devices Inc., Pullman, WA, USA). Soil ψ was taken on the same days as the gas exchange measurements, with five replications per site per day.

Redox potential

Redox potential (E_h) was determined with brightened platinum electrodes (Faulkner et al., 1989) inserted in the soil to 8 cm. Prior to usage, the electrodes were checked for accuracy in pH 7 buffer with quinhydrone. Electrodes were allowed to equilibrate for at least 1 h prior to E_h measurement. E_h was calculated by adding the potential of a standard calomel reference (+244 mV) to the reading. E_h was taken on the same days as the gas exchange measurements, with five replications per site per day.

RWC and chlorophyll content.

Leaf relative water content (RWC) was measured on mature leaves by determining fresh mass (FM), then placing the leaf in saturating conditions for 24 h. After determining turgid mass (TM), leaves were oven-dried for 72 h at 70 °C and re-weighed to determine dry mass (DM). Leaf relative water content was calculated as (FM – DM)/(TM – DM) and expressed as a percentage.

Chlorophyll content was determined with a hand-held chlorophyll absorbance meter (Opti Sciences, Tyngsboro, MA, USA) following instructions recommended by the manufacturers. Hand-held non-invasive chlorophyll meters, like the CCM-200, have been found to provide a reliable estimate of leaf chlorophyll (Richardson et al., 2002). Measurements of RWC and chlorophyll content were taken on the same days as the gas exchange data, with 25 replications per site per day over 6 d.

Statistical analyses

Data, which satisfied the assumption of ANOVA, were subjected to unpaired t-tests to detect for differences between the two sites. Other data were subjected to one-way analyses of variance and Tukey's compromise test (P < 0·05) using GraphPad Instat Version 3.00 (Mustek). Regression of rates of photosynthesis over variation in leaf conductance was performed in SIGMAPLOT 6.1 (SPSS Science) using an exponential function.

RESULTS

Soils

The concentration of total soil cations and electrical conductivity of the soil saturation extract were higher in the dwarf site by 70 % and 63 %, respectively, compared with the fringe site. In the dwarf site, concentrations of Na⁺, K⁺, Ca²⁺ and Mg²⁺ were higher than in the fringe site by 70 %, 25 %, 39 % and 97 %, respectively (Fig. 1). Soil P was the only element that was significantly lower (by 29 %) in the dwarf zone. Concentrations of the microelements Zn²⁺, Mn²⁺ and Cu²⁺ were also higher in the dwarf site by 89 %, 44 % and 68 %, respectively (Fig. 1). Soil pH at both sites was alkaline, being 7·63 in the fringe zone and 8·39 in the in the dwarf zone. There were no differences in bulk density or redox potential between the fringe and dwarf sites. Bulk density was 1·68 ± 0·03 g cm^–3 in the fringe and 1·55 ± 0·04 g cm^–3 in the dwarf site, while redox potentials were –40·4 ± 15·8 mV and –26·0 ± 15·6 mV, respectively.

Fig. 1.

Total cations (TC), electrical conductivity (EC) and concentrations of Na, K, Ca, Mg, P, Zn, Mn and Cu in soils from fringe and dwarf mangrove sites. Means ± s.e. are given; n = 7.

Plant analyses

The predominant ion in the leaves was Na⁺, concentrations being 24 % higher in dwarf than in fringe mangroves. By contrast, concentrations of K⁺, Ca²⁺, Mg²⁺and P were significantly lower in the leaves of dwarf mangroves by 49 %, 37 %, 36 % and 49 %, respectively. Nitrogen concentration in leaves of dwarf mangroves was 20 % higher than those in the fringe zone (Fig. 2). Leaf concentrations of Zn²⁺, Cu²⁺ and Fe³⁺ in fringe and dwarf mangrove leaves were 20 ± 2 and 35 ± 13, 14·7 ± 2·7 and 14·6 ± 2·2 and 284 ± 6 and 367 ± 20 mg kg^–1, respectively, with no differences between sites. The concentration of Mn²⁺ in dwarf leaves was 27 % lower than in the fringe leaves.

Fig. 2.

Concentrations of Na, K, Ca, Mg, N and P in mature leaves of fringe and dwarf Avicennia marina. Means ± s.e. are given; n = 5.

In the fringe mangroves, average CO₂ exchange was 11·86 μmol m⁻² s⁻¹, while in the dwarf site this was significantly reduced by 48 %. In dwarf plants, leaf conductance was consistently lower than those in the fringe zone, suggesting that light-saturated photosynthesis is conductance limited (Fig. 3). Quantum yield of PSII electron transport (yield) and ETR through PSII were significantly reduced by 39 % and 37 %, respectively, in the dwarf site compared with the fringe site. Intrinsic PSII efficiency (F_v/F_m) was lower by 6 % in the dwarf site compared with the fringe site (Fig. 4). In the fringe site, F_v/F_m decreased from 0·7940 at dawn to 0·7688 at midday, but recovered overnight. The leaf chlorophyll content index (CCI) values were higher in fringe mangroves and significantly reduced by 24 % in dwarf mangroves (Fig. 4).

Fig. 3.

Relationship between light-saturated CO₂ exchange and leaf conductance in fringe (closed circles) and dwarf (open circles) Avicennia marina. Means ± s.e. are given, n = 180. Measurements were taken at saturating light (1200–1500 μmol m⁻² s⁻¹) and at 28−30 °C.

Fig. 4.

Photosystem II quantum yield, electron transport rate through PSII and photochemical efficiency of PSII (F_v/F_m) and chlorophyll content index in fringe and dwarf Avicennia marina. Means ± s.e. are given; n = 180 for yield and ETR, 30 for F_v/F_m and 150 for CCI. The CCI index is proportional to the amount of chlorophyll in the leaf.

Soil ψ decreased progressively with increasing elevation and was −3·07 MPa in the fringe site, −7·44 MPa in the dwarf site (Fig. 5) and −10·27 MPa in the barren zone.

Fig. 5.

Soil water potential, midday minimum xylem water potential, % RWC and dry mass : turgid mass ratio in fringe and dwarf Avicennia marina. Means ± s.e. are given; n = 30 for soil ψ, 9 for xylem ψ and 125 for RWC and DM/TM.

Midday minimum xylem ψ was −4·5 MPa in the fringe site and −6·5 MPa in the dwarf site (Fig. 5). Leaf RWC was 96 % in fringe mangroves and significantly lower by 10 % in dwarf mangroves. Dry mass : turgid mass was higher by 9 % in dwarf leaves (Fig. 5)

DISCUSSION

In this study, seasonal or tidal effects were minimized by collection of measurements and samples at different times of the year and at different tidal cycles. Moreover, differences between sites were consistent irrespective of tidal cycle or time of data collection. In the fringe zone, regular tidal inundation, twice daily, flushes out excess solutes in the soil and maintains a soil ψ (−3·07 MPa) that is close to that of seawater (−2·5 MPa). By contrast, hypersalinity conditions exist in the dwarf site. Infrequent tidal coverage (only at high spring tides) and high evapotranspiration in the dwarf site contribute to higher concentrations of ions and higher electrical conductivity in the soil than in the fringe zone. The more alkaline soil pH in the dwarf zone could have influenced solubility of ions such as P, Fe³⁺ and Mn²⁺ and, therefore, their concentrations in the soil and plants.

In the dwarf site, high soil concentrations of inorganic ions contribute to extremely low soil water potentials that produce a series of effects similar to drought stress. Midday minimum xylem ψ in the fringe zone was consistently lower than soil ψ suggesting a positive plant water balance. In the dwarf site, xylem ψ at midday was higher than soil ψ and, together with the hydraulic conductance of the soil–plant system, would result in insufficient water transport to the shoot. Soils in both sites were always moist and there was little variation in soil ψ within replications. Preliminary data indicated that soil ψ was similar up to a depth of about 30 cm. Plants in the dwarf site experience a negative water balance during midday and probably recover overnight, after rainfall and during inundation at high spring tides. Several workers have demonstrated that mangroves and, especially salt secreters such as Avicennia species, are able to regulate xylem sap osmotic potentials to levels much lower than soil osmotic potential (Paliyavuth et al., 2004; Lopez-Portillo et al., 2005) primarily through uptake of Na⁺ and Cl⁻ (Ball, 1996). Reported xylem sap concentrations for mangroves range from 0·5 to 380 mm NaCl (Scholander et al., 1962; Sobrado, 2002). This adaptation enables species such as Avicennia to maintain a positive water balance under a broad range of salinities (Sobrado, 2002). In A. germinans, hydraulic conductivity was shown to decrease with high xylem sap salinity (Lopez-Portillo et al., 2005) consistent with the concept of swelling and shrinkage of pectin-based hydrogels in the xylem pit membranes (Zwieniecki et al., 2001). At very high salinities, such as those that exist in the dwarf site, hydrogels may play an important role in maximizing hydraulic conductivity during the day and growth at night as a result of higher turgor in meristematic cells (Lopez-Portillo et al., 2005).

The results indicate that dwarf mangroves are probably under severe water stress and may access water from microsites in the soil that are less negative in water potential. The frequent and sometimes extensive die back of branches, as well as the smaller and thicker leaves of dwarf plants (Hiralal et al., 2003) indicate a low capacity for water transport and reduced turgor, cell expansion and growth. High soil salinities impair nutrient acquisition in salt-marsh species (Flowers et al., 1977; Greenway and Munns, 1980) and mangroves (Ball and Farquhar, 1984; Popp et al., 1985). Preferential uptake of Na⁺ and significantly reduced acquisition of K⁺, Ca²⁺ and Mg²⁺ at high salinities are well documented in many halophytes including mangroves (Popp et al., 1985). High concentration of Na⁺ in the substrate impairs K⁺ acquisition by roots due to physico-chemical similarities between these two elements and leads to K⁺ deficiency (Marschner, 1995) and a low K⁺ : Na⁺ ratio in plant tissues, which is one of the key elements in salinity tolerance (Yeo, 1998; Maathuis and Amtmann, 1999). Altered nutrient acquisition is a consequence of high substrate salinity and would be expected to impair metabolism as well as growth of dwarf mangroves.

Lack of differences in the concentration of microelements between fringe and dwarf leaves, except for Mn²⁺, may be due to the use of mature leaves for the analyses. Younger leaves may have yielded different results, especially for those elements with low mobility in the xylem (Epstein, 1972). Higher concentrations of N in dwarf mangrove leaves may be due to differences in contribution of tides and also to upland influence.

The data suggested that the dwarf site may be P-limited, soil and leaf P concentrations being 29 % and 49 % lower, respectively, than in the fringe zone. P limitation in the dwarf site may be attributed to infrequent tidal influence and impaired plant nutrient balance. Low P availability in the dwarf site may also limit microbial processes such as rates of litter decomposition and nutrient retention in the rhizosphere and further impair nutrient acquisition (Amador and Jones, 1993).

The observed leaf P deficiency in dwarf leaves may be a consequence of impaired metabolism due to altered K⁺ : Na⁺ ratios and to low soil P availability. In other parts of the world, growth of dwarf plants of R. mangle has been enhanced by addition of P (McKee et al., 2002; Lovelock et al., 2004) suggesting that it is the lack of P that limits growth. Nutrient enrichment studies on dwarf R. mangle in Panama (Lovelock et al., 2004) and Belize (Feller, 1995; McKee et al., 2002) suggested that growth was partially limited by N and P deficiency. Shoot growth of dwarf plants increased 10-fold with P and 2-fold with N fertilization primarily due to increases in stem hydraulic conductance, which increased 6-fold by P and 2·5-fold with N enrichment (Lovelock et al., 2004).

Isotopic fractionation studies on dwarf R. mangle in Belize suggested that growth was limited by the interaction of external supply, internal demand and plant ability to acquire N and P under different hydro-edaphic conditions (McKee et al., 2002). In dwarf plants, P limitation may reduce hydraulic conductance through reduction in cell expansion (Grossman and Takahashi, 2001), decreases in dimensions of xylem vessels, porosity of xylem pit membranes and hydrogel content (Tyree and Ewers, 1996; Comstock and Sperry, 2000; Lopez-Portillo et al., 2005) and the number and functioning of aquaporins within the root plasma membranes (Carvajal et al., 1996; Clarkson et al., 2000).

High substrate salinities in the dwarf site caused a build-up of Na⁺ in the leaves, impaired nutrient acquision and probably reduced hydraulic conductivity and water transport to the leaves. These changes would be expected to have adverse effects on various metabolic processes such as CO₂ exchange. In this study, reductions in mean CO₂ exchange and mean leaf conductance between fringe and dwarf mangroves were 48 % and 44 %, respectively, which are similar to those (43 % and 57 %) reported for A. marina at seawater and hypersalinity (60 ‰) in Australia (Sobrado and Ball, 1999). Photosynthetic rates were correlated with stomatal conductance (Fig. 3; R² = 0·62, P < 0·001) with mean conductance being significantly lower in the dwarf plants.

Parallel decreases in CO₂ exchange and conductance as a consequence of low water availability, reduced RWC and chlorophyll content, suggest that RUBP regeneration and ATP synthesis would be impaired in dwarf mangroves (Lawlor and Cornic, 2002). In dwarf mangroves, the smaller, more xeromorphic leaves with thicker cuticles (Hiralal et al., 2003) may have a lower capacity for CO₂ diffusion to the photosynthetic tissues. Moreover, others have shown that photosynthetic capacity decreases at high salinity (Sobrado and Ball, 1999; Tuffers et al., 2001). The gas exchange data suggest that spatial variation in environmental stresses across a productivity gradient reduces photosynthetic performance via stomatal conductance and/or carboxylation.

In the dwarf site, decreases in the efficiency of open PSII reaction centres in the light, ETR through PSII and photochemical efficiency of PSII in dark-adapted leaves support the gas exchange data that CO₂ exchange was reduced at this site. In addition to high substrate salinity, dwarf mangroves also have to contend with high radiation loading because trees are isolated and lack a vegetative canopy.

Photoinhibition, the light-dependent reduction of the intrinsic quantum yield of photosynthesis and a loss of PSII activity (Osmond, 1994) would be exacerbated at high light and high salinities in dwarf leaves. Persistent depression in F_v/F_m in A. marina seedlings at high salinities have been reported (Bjorkman et al., 1988). In the fringe mangroves, midday reductions in the efficiency of PSII were recovered at dawn, as reported for other mangroves (Cheeseman et al., 1997; Sobrado and Ball, 1999) and are probably the result of protective down-regulation (Osmond, 1994). In dwarf mangrove leaves, following high irradiance with high temperatures and no rainfall or tidal coverage, midday reductions in the efficiency of PSII were not completely recovered at dawn suggesting chronic photoinhibition (Osmond, 1994). Additional adaptive strategies of dwarf A. marina to reduce photon capture for energy transduction and use include lower leaf chlorophyll (Fig. 4) and steeper leaf angles (Bjorkman et al., 1988; Tuffers et al., 1999; Christian, 2005).

This study has demonstrated that dwarfing in mangroves is a complex phenomenon influenced by a variety of hydro-edaphic conditions, as suggested for R. mangle in Panama (Lovelock et al., 2004) and in Belize (McKee et al., 2002). Infrequent tidal coverage and high evapotranspiration contribute to extremely low soil water potentials that impose severe plant water stress. Moreover, high substrate concentrations of ions, specifically Na⁺, disrupt nutrient uptake and cause ion imbalance in leaves. This study also indicated that P limitation may be implicated in dwarf mangrove production, as both soil and leaf levels of this element were significantly lower in dwarf than in the fringe zone. These findings tend to support those of others (McKee et al., 2002; Lovelock et al., 2004) that the primary nutrient limiting growth of dwarf scrub R. mangle in Panama, Belize and Florida is P. Additional research is needed to test and further define these conclusions.